Claude and LinkedIn teaming up to bring you hot leads on autopilot

I sell LinkedIn Automation APIs. I cold emailed 319 VC deal sourcing platforms that nobody else was targeting. Here's what happened. My colleague @AshwiniNK21 runs outbound for my SaaS. A few months ago she came to me with a list and a targeting angle I hadn't seen before. It was seriously, f***ing brilliant. The targeting behind that deal was a vertical we'd never touched before. 319 VC deal sourcing platforms across 22 countries. Deal sourcing runs on LinkedIn, that's where founders announce rounds, where portfolio companies post updates, where deal flow actually lives. But none of the deal sourcing platforms these funds use connect to LinkedIn natively (they just behave like a CRM) That gap had been sitting there completely untouched. What I hadn't appreciated until she explained it is how she actually finds verticals like this. She starts with Discolike, feeds it a company that looks like an ideal customer, and gets lookalikes back. Half of them make you go "who?" but the AI sees patterns humans don't. Then she cross-checks on Crunchbase to confirm whether companies are actually alive or just a LinkedIn page with no employees, enriches through Clay to find the right decision-makers, and verifies every email with BetterContact. The list after all of that is the actual list. One of the best leads that came through from there was an ex-venture capitalist based out of New Zealand turned founder who was making software for VC firms to source deals. He replied: "We don't do that, but I am interested." Two emails in, he booked a call, and we're now helping him set up LinkedIn outreach from scratch. The reason I am writing this full post is that cold email really works if you have a unique angle and a persona that not enough people reach out to, and a data source that is relatively untouched. If you use typical database tools like Apollo or LinkedIn Sales Navigator, you are not going very far because most of the people on those databases get thousands of cold emails every month.

Rule No 1. Never scroll past a Prerona Basu article.

just don't share this with any CMO you know. will piss them off

Every acquisition offer has two numbers. The headline number. And the real number. Most founders only read the headline. In 2022 a PE firm offered to buy SalesRobot for $1,675,000. We were at $550K ARR, bootstrapped, five people from India. I'd started the company after getting fired during the pandemic with zero savings. $1,675,000 felt like the exit. Then I actually read the term sheet. Here's how the $1,675,000 broke down: • $75K at closing for the codebase. • $575K for 35% equity. • $1,025,000 in installments over 24 months - contingent on financing they didn't yet have. • Plus a 12-month employment lock-in. Guaranteed cash at closing: $650,000. The other million was a bet on a stranger's ability to get a loan. That's the gap nobody talks about. The headline number is what gets announced. The real number is what lands in your account if everything goes exactly as planned. I walked away on gut feel. Didn't do the math properly until three years later. We're at $75K MRR now. At a conservative 3x multiple the business is worth roughly $2.7M. I kept 100% of it. No non-compete. No boss. No bet on someone else's bank loan. Three things to read before you sign anything: The installment terms - what triggers them and what voids them. The financing conditions - who carries the risk if the buyer can't secure funding. The non-compete - what you can't build and for how long. The headline number is almost never the real number.

I built a Fractional CMO using Claude Opus 4.8. 7 specialist skills that replace a $12-15K/mo CMO hire And run your entire marketing engine from strategy to distribution. The Claude setup is ready in minutes and gets you results from day 1 You get 7 specialists working in sync: 1. Positioning Researcher → analyzes ICP, competitors, market gaps 2. Content Strategist → pulls search demand, audience signals, competitor content gaps to build editorial calendar 3. Copywriter → drafts LinkedIn posts, blog intros, newsletters, and landing page copy in your voice 4. Repurposing Engine → takes one long-form piece and turns it into 8-10 assets across LinkedIn, X, email, blog, and short-form video scripts 5. SEO & AEO Optimizer → handles keyword mapping, metadata, heading structure, and citation signals that get your content on AI search engines 6. Distribution Planner → maps every piece to the right channel, format, and posting window so nothing sits in a Google Doc collecting dust 7. Performance Analyst → tracks what's actually driving pipeline, flags what to kill, and recommends where to double down What you get when all 7 work in sync: → A positioning doc you can actually use → Content that targets real search demand → One piece of content multiplied across every channel → SEO and AI-engine visibility baked in from the start → A distribution system that runs whether you're online or not → Clear signal on what's working so you stop funding dead channels You can set it up today. Bookmark this post so you don't lose it. Comment "CMO" and I'll send it. ♻️ Repost for priority access.

Can’t wait to see how this turns out!

Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors. Available today at the same price.

Right now the best way to keep a secret is to post about it on LinkedIn. No one will get to see it.

I disappointed my mother again… She heard me talk about Hermes and showed me a picture of a bag she wanted. Then I told her Hermes was an AI agent. She put the phone down and walked away. Now how do I tell her my Hermes actually remembers. Her’s doesn't. I dont want to bug her with more AI talk but you need to hear me out. If you're running LinkedIn outreach right know the difference between Claude vs Hermes. Claude is brilliant. I use it every day. But every time you close the tab, it forgets you completely. ICP. Tone. Sequences. Who dis? So every session starts the same way: Paste context, Re-explain your offer, Your prospect's situation, Then finally do the actual work. Hermes runs on a server. After every task it writes its own notes: What it did, What worked, What to do faster next time. Day 1: You explain your workflow. Day 10: It already knows it. Imagine what that means when you’re running 10 client accounts. While Claude will bleed you in tokens as you're rebuilding context for every client every single session. Hermes pays for itself in week one. I built 5 Hermes skill files specifically for LinkedIn outreach so your agent hits the ground running from day one: Cold DM Writer Reply Classifier Follow-up Sequencer Objection Handler Lead Scorer All 5 are pre-built. Drop straight into your Hermes skills folder. Comment “HERMES” and I'll send them over. I mean the skills btw. Not the bag. Yet

Our head of social fixed her LinkedIn About section. Got 2 job offers the next week. She wasn't looking. Wasn't applying. Just cleaned up the first thing anyone sees when they land on her profile. @HelloPrerona didn't do anything complicated. No algorithm hacks. No posting at 8 AM on Tuesdays because someone on YouTube said so. Just a sharper bio that actually said what she does and who she helps. That's it. We built EinsteIn because this keeps happening. People post for months. Decent content. Okay engagement. Zero leads. Then they fix their bio and everything changes. Free tool. Takes minutes. No credit card. Comment "ROAST" and I'll send you the link.

Come home I'll make you garlic ice-cream tacos. Don't want them? Why? After reading your emails, I thought you liked unappetizing, barf-worthy things. Made my point. Now LISTEN This is why you’re not getting replies to your cold emails. You’re → Congratulating someone on a job change 30 days after it happened.  They've moved on. So has the relevance. → Opening with "I noticed you recently posted about…"  They know you didn't notice. An AI noticed. → Running an 8-step sequence.  By step 4 you're in spam. By step 8 you're blocked. → Pulling the same enrichment data as every other tool targeting your ICP.  Funding rounds, recent LinkedIn posts, company size. Commoditized data produces commoditized messages. → Letting the AI decide what to personalize.  It will always pick the most obvious signal. Which is also the most tired one. → Automating everything up to the reply Then doing reply management manually. You saved 4 hours on prospecting and lost 6 hours reading inboxes. The irony: most of this was correct advice 3 years ago.  The playbook rotted.  Nobody updated it. The last one is what we built Kuron to fix. An agent that doesn't stop when someone replies.  It classifies the reply, handles the objection, and books the meeting directly to calendar. No human in the loop.  Including at the reply stage. → Join the waitlist at kuron.ai PS: DM me "garlic ice-cream tacos" for the full recipe.

A UK agency built a $50K/month business on top of a tool we were charging them $400/month for. They came in a couple years ago with 5 seats. Standard white label deal. Their branding, their domain, their clients never knowing SalesRobot existed. We charged them $100/month white label fee plus $59/seat. Roughly $400 a month total. They bundled us with GoHighLevel and sold the whole stack to their clients as a complete outbound system. Every new client they onboarded got LinkedIn automation baked into the package. They grew quietly. We did nothing except keep the product running. 5 seats became 80 seats over the next 12 months. At some point they were doing $50,000 a month in revenue just from the SalesRobot white label. We were making $400. The mistake wasn't that we undercharged. The mistake was that nobody asked what they were building before agreeing to a price. We assumed they were running a few campaigns for a handful of clients. Standard use case. We priced accordingly. Their use case was a $50K/month business built on our infrastructure. When someone asks for a white label deal, the first question shouldn't be "what does this cost us to provide." It should be "what is this worth to them." What would they pay to build it themselves? What would they lose if they had to go without it? What are they planning to charge their clients and How many clients do they expect? That one conversation changes the entire pricing discussion. The pattern has repeated since. Agencies building real businesses on top of a product priced like a SaaS subscription. The tool costs don't change. The value to the agency is completely different. How do you guys think about pricing for custom deals like this?

The scariest moment in a bootstrapped SaaS isn't year one with no revenue. It's year five. With $1M ARR. And four channels dying at the same time. For the last 90 days I've had a front row seat to something nobody warns you about. SEO was our most consistent channel for two years. Then AI overviews quietly ate our traffic. No dramatic crash. Just fewer clicks every week until the trend became impossible to ignore. LinkedIn reach started dropping. Same posting frequency. Same quality. 3.3 million impressions across 2025. Noticeably less distribution now. Outbound reply rates fell. Conversion rates fell. The channel still works. Just less, for the same effort. Cold calling: Hired someone specifically to plug the gap. Still nothing yet. The frustrating part isn't that one channel faded. It's that all four weakened in the same window. The honest diagnosis our founder landed on: we got very good at capturing existing demand. SEO. Outbound. LinkedIn content. We intercepted people who already knew they had a problem. We never built the thing that made customers talk about us unprompted. No word of mouth. No engine underneath the engine. So now I'm watching him go back to the drawing board in real time. Meta ads. Reddit. A second product built on Claude Code - an AI system that runs outreach the way a strong agency would. I don't know if it's the right bet either. But there's something clarifying about watching a founder refuse to squeeze harder on things that are already fading. Sometimes the bravest thing isn't pushing through. It's admitting the foundation needs rebuilding.

Claude can watch your competitor's YouTube videos and tell you exactly How they're positioning against you. What features they led with. What they buried at 9 min 24 sec. The exact moment they positioned against you without saying your name. It can actually WATCH the video frame by frame and not just read the transcript. Isn’t that freakin amazing!?! The system inside Claude works like this: → paste any YouTube link → Claude pulls the actual video → splits it into frames across the full runtime → reads the visuals and the audio together → hands you a structured breakdown in under 2 minutes Last week I fed it 5 competitor product demos back to back. In 10 minutes I had: → the exact features each one led with in their first 30 seconds → what they showed on screen vs. what they actually said → where they spent the most time (tells you what they think their edge is) → the objections they preemptively handled in the demo (tells you what they're losing deals on) That last one is worth reading twice. Claude pulls out every pain point they mentioned so you can mirror their language in your first message. ❌ no expensive video API ❌ no extra tools to buy ❌ no watching 4 hours of competitor content with a notepad ✅ one SKILL running inside Claude (you just need a Claude Pro account) I put the full setup walkthrough into a playbook, if you want it: Comment WATCH And I’ll DM you the link. Simple. Bookmark this post so you can come back to it before the next quarter.

I got fired during COVID with 0 savings. The idea I actually wanted to build cost $75,000 just to start. The “backup” idea is now at $1.1M ARR. In 2020 I was building a cold email tool. Lemlist was growing fast and the market felt obvious. Warm channel, high intent, clear ICP. But then I found out Gmail had raised API access fees to $75,000. That was it. Idea over. I had no savings, no runway, no fallback. I needed something I could actually build. LinkedIn outreach had the same problem I was trying to solve. Warmer channel and a lower barrier. And nobody had built a “reliable” cloud-based tool for it yet. It felt like a backup plan at the time, not a real idea. So I started building. The first breakthrough came from a message I sent to a Facebook group for users of Dux-Soup (a LinkedIn automation tool I knew was terrible). 4,000 people were in that group, so I scraped them all with PhantomBuster and sent one message to every single one: "Hey, saw you in the Dux-Soup group. Are you still using it for LinkedIn lead gen?" At 3 am on a fine night in May 2020, a guy named Connor replied. He ran a lead gen agency out of Manchester. 80 LinkedIn accounts. A call center in Zimbabwe with two people logging in every morning, one by one, just to check if each client's account was still connected. He said: "If you can build something reliable in the cloud, I'll move immediately." He came in with 50 seats, and paid me $4,000 MRR. And this was before the product was worth talking about. I'm based in India, so before anyone does the math assuming wild numbers, the comparison is against Indian tech salaries, not Western ones. But $1M ARR bootstrapped with 0 VC money after getting fired with 0 savings still feels surreal to say out loud. Yes, it did take almost 6 years. But the $75,000 API fee that made me drop my first idea pushed me to a channel that turned out to have better fundamentals. I don't want to get all "everything happens for a reason" wishy-washy nonsense on you, but sometimes it does. So take all the setbacks you face in your startup journey as something that might lead you down a better path 🙂

My LinkedIn automation SaaS hit $1M ARR this year after 5 years. Fully bootstrapped, India based team, 0 VC money. And for the last 90 days, everything that got me there has started underperforming. Here's what’s been happening: SEO was our most consistent channel for two years. Then AI overviews started changing search behaviour and Google slowly stopped sending us traffic. No clear moment where it broke. Just fewer clicks every week until the trend became impossible to ignore. LinkedIn posting worked well through most of 2025. 5 people on our team posting consistently, going viral regularly, 3.3 million impressions across the year. Then reach started falling. Same posting frequency, same quality, noticeably lower distribution now. Outbound was always the foundation. It's what got us from $0 to $300K ARR in the first year. But reply rates and conversion rates have both been declining. The channel still works. Just less than it used to for the same amount of effort. Then we hired a cold caller specifically to help fill the gap. That channel still hasn't produced anything yet. The frustrating part isn't that one channel declined. Channels fade all the time. The frustrating part is that all 4 weakened in the same window, which usually means the issue is bigger than channel execution. We have almost no word of mouth. That's the honest diagnosis. Most of our growth came from intercepting people who already knew they had a problem. SEO. Outbound. LinkedIn content. We got good at capturing existing demand. We never built the thing that made customers naturally talk about us to other people. And when all the demand capture channels weaken together, there's nothing underneath them. So right now I'm back at the drawing board. I'm considering Meta ads. I started taking Reddit seriously this year because more buyers are discovering tools through AI search results instead of Google. And I'm building a second SaaS because 5 years of watching customers use the first one exposed a gap it couldn't solve. The second product is an AI system that runs outreach the way a strong agency would. Built on Claude Code. Trained on real campaign data and workflows from people actually doing the work every day. I don't know yet if it's the right bet. But when all your existing channels start weakening at the same time, building something with a different foundation felt more honest than squeezing harder on things that are already fading. Still figuring it out.